摘要
在分析红外场景模型的基础上提出了用自适应 Wiener 滤波器估计起伏背景、自适应门限分割、基于邻域管道目标检测的小目标检测方法。采用连续采集的红外图像序列进行了实验,仿真了不同信噪比(SNR)条件下的目标并给出了目标检测结果及其分析。结果表明,算法能够从信噪比大于 2.0的图像序列中检测出目标轨迹。
In this paper, IR scene model is first introduced, then an algorithm based on the clutter background suppression with adaptive Wiener filter is put foreword, and the small targets are detected with adaptive threshholding and neighborhood-pipeline detection method. Experiments are done with long-range IR image sequences. The results show that this algorithm can detect small targets within IR image sequence with SNR higher than 2.0.
出处
《红外技术》
CSCD
北大核心
2004年第6期54-57,61,共5页
Infrared Technology
